Mako hSDM covariate explore (ROMS and CMEM domains)

Author

Emily Nazario

Published

January 31, 2024

For each of the figures below, I have visualized what the covariate values are for the CRW PA positions and observed positions. Some values were extracted using the ROMS data, and thus limiting the position range to this domain. I have also explored the covariate values using the full range of the available positions using covariate files from Copernicus (CMEM). Please note that the lists of covariates selected for each data source do vary depending on what was recommended and available.

The covariate values explored on this page are all from the surface, and thus depth is not yet considered in this analysis.

Lastly, for each collection of a shark’s observed positions, there are 32 replicate PA CRW tracks for the CMEM domain and 16 replicate tracks for the ROMS domain.

Covariate summary plots for all sharks

CMEM figures

Covariate selection justification

The figures below only represent covariate values for observed positions. PA positions were omitted from this analysis due to the quantity.

Covar vs. covar scatter

Below are some comparisons of the covariate values relative to each other for the observed positions (covariate values at PA positions are excluded).

ROMS figures

Covar vs. covar scatter

Select shark analyses

I randomly selected three tag IDs to include for this next analysis.

set.seed(1004)
sample(1:23, 3) #randomly select 3 sharks: 11, 22, and 1
[1] 11 22  1
#filter to only keep the tag IDs of the above three sharks.
cmem_dat_sub <- cmem_dat %>% filter(tag == "41770" | tag == "68285" | tag == "78151")
cmem_obsv_sub <- cmem_dat_sub %>% filter(PA == 0)

I then randomly select which repetition number I want to include as a representative for this series of analyses. I am left with a subset of the data which will be used for the Select Shark Analyses portion of this document.

sample(1:32, 1) 
[1] 13
cmem_rep_sub = NULL
for(i in 1:length(unique(cmem_dat_sub$tag))){
  curr_ID <- unique(cmem_dat_sub$tag)[i]
  temp_df <- cmem_dat_sub[cmem_dat_sub$tag %in% curr_ID,]
  temp_df <- temp_df %>% group_by(rep)
  
  curr_rep <- unique(temp_df$rep)[13] #number randomly selected above
  temp_df2 <- temp_df[temp_df$rep %in% curr_rep,]
  
  cmem_rep_sub <- rbind(cmem_rep_sub, temp_df2)
}

cmem_dat_sub2 <- rbind(cmem_obsv_sub, cmem_rep_sub)

CMEM AGI

These figures track how DO changes through a shark’s track for both its observed positions (‘0’) and PA positions (‘1’).

CMEM DO

These figures track how DO changes through a shark’s track for both its observed positions (‘0’) and PA positions (‘1’).

CMEM Temp.

These figures track how temp. changes through a shark’s track for both its observed positions (‘0’) and PA positions (‘1’).

ROMS DO

These figures track how DO changes through a shark’s track for both its observed positions (‘0’) and PA positions (‘1’).

ROMS Temp.

These figures track how temp. changes through a shark’s track for both its observed positions (‘0’) and PA positions (‘1’).

Maps

Plot DO and temp. values over space for a few representative dates for sanity check of mapped values. Randomly selected three days (day)

CMEM

ROMS